✨ 搜索支持区分繁简选项 Add Simplified/Traditional Chinese sensitivity option for search

新增搜索配置 hanSensitive（区分繁简，默认开启 = 现状行为，命名与语义
向 caseSensitive 对齐）。关闭后简体与繁体可互相检索。

- FTS 层：关闭区分繁简时，blocks_fts / blocks_fts_case_insensitive 建表
  追加 han_insensitive 分词器参数（依赖 88250/go-sqlite3 配套补丁），
  繁体在分词时按单字折叠为简体，token 偏移仍指向原文；
- 切换选项复用区分大小写现有的 FullReindex 流程；
- 高亮：EncloseHighlighting 在关闭区分繁简时将关键字逐字符展开为
  繁简等价字符类（如 诗经 -> [诗詩][经經]），与分词器使用同一份
  OpenCC TSCharacters 映射数据；
- 兼容：旧版本 conf.json 缺 hanSensitive 字段时按既往行为置为开启；
  setSearch 请求未携带该字段（现行前端/第三方调用）时保持当前值，
  避免被零值意外翻转并触发重建索引；
- 遵照配置界面重写计划（#17601），本次不改动设置界面与 i18n，仅添加
  配置数据结构字段与前端类型定义（config.d.ts）。

范围说明：本次仅覆盖块全文搜索（关键字/查询语法）的命中与高亮；
历史/资源文件搜索、数据库属性过滤、虚拟引用、查找替换的替换动作
不在本次范围内（替换保持字节精确以避免误改原文）。

Fixes #13304

Author: fujingzhai

app/src/types/config.d.ts

@@ -1336,6 +1336,10 @@ declare namespace Config {
          * Whether to search embedded blocks
          */
         embedBlock: boolean;
+        /**
+         * Whether to distinguish between Simplified and Traditional Chinese characters when searching
+         */
+        hanSensitive: boolean;
         /**
          * Whether to search heading blocks
          */

kernel/api/setting.go

@@ -501,11 +501,17 @@ func setSearch(c *gin.Context) {
 		return
 	}
 
+	if _, exists := arg["hanSensitive"]; !exists {
+		// 兼容未携带该字段的旧版前端/第三方调用：保持当前值，避免被零值意外关闭并触发重建索引
+		s.HanSensitive = model.Conf.Search.HanSensitive
+	}
+
 	if 32 > s.Limit {
 		s.Limit = 32
 	}
 
 	oldCaseSensitive := model.Conf.Search.CaseSensitive
+	oldHanSensitive := model.Conf.Search.HanSensitive
 	oldIndexAssetPath := model.Conf.Search.IndexAssetPath
 
 	oldVirtualRefName := model.Conf.Search.VirtualRefName
@@ -517,9 +523,10 @@ func setSearch(c *gin.Context) {
 	model.Conf.Save()
 
 	sql.SetCaseSensitive(s.CaseSensitive)
+	sql.SetHanSensitive(s.HanSensitive)
 	sql.SetIndexAssetPath(s.IndexAssetPath)
 
-	if needFullReindex := s.CaseSensitive != oldCaseSensitive || s.IndexAssetPath != oldIndexAssetPath; needFullReindex {
+	if needFullReindex := s.CaseSensitive != oldCaseSensitive || s.HanSensitive != oldHanSensitive || s.IndexAssetPath != oldIndexAssetPath; needFullReindex {
 		model.FullReindex(false)
 	}
 

kernel/cli/cmd/root.go

@@ -87,6 +87,7 @@ var rootCmd = &cobra.Command{
 		logging.SetLogToStdout(false)
 
 		model.InitConf()
+		sql.SetHanSensitive(model.Conf.Search.HanSensitive) // 须在 InitDatabase 之前设置：FTS 建表的分词器参数依赖该选项
 		sql.InitDatabase(false)
 		sql.InitHistoryDatabase(false)
 		sql.InitAssetContentDatabase(false)

kernel/cli/cmd/serve.go

@@ -32,8 +32,8 @@ import (
 )
 
 var serveCmd = &cobra.Command{
-	Use:   "serve",
-	Short: "Start kernel HTTP server",
+	Use:                "serve",
+	Short:              "Start kernel HTTP server",
 	FParseErrWhitelist: cobra.FParseErrWhitelist{UnknownFlags: true},
 	PersistentPreRunE: func(cmd *cobra.Command, args []string) error {
 		return nil // bypass root's init — Boot() handles it
@@ -67,6 +67,7 @@ var serveCmd = &cobra.Command{
 		model.InitConf()
 		go server.Serve(false, model.Conf.CookieKey)
 		model.InitAppearance()
+		sql.SetHanSensitive(model.Conf.Search.HanSensitive) // 须在 InitDatabase 之前设置：FTS 建表的分词器参数依赖该选项
 		sql.InitDatabase(false)
 		sql.InitHistoryDatabase(false)
 		sql.InitAssetContentDatabase(false)

kernel/conf/search.go

@@ -46,6 +46,7 @@ type Search struct {
 
 	Limit         int  `json:"limit"`
 	CaseSensitive bool `json:"caseSensitive"`
+	HanSensitive  bool `json:"hanSensitive"` // 区分繁简：默认开启（与既往行为一致）；关闭后全文搜索不区分简体/繁体中文字形
 
 	Name  bool `json:"name"`
 	Alias bool `json:"alias"`
@@ -89,6 +90,7 @@ func NewSearch() *Search {
 
 		Limit:         64,
 		CaseSensitive: false,
+		HanSensitive:  true,
 
 		Name:  true,
 		Alias: true,

kernel/harmony/kernel.go

@@ -54,6 +54,7 @@ func StartKernel(container, appDir, workspaceBaseDir, timezoneID, localIPs, lang
 	go server.Serve(false, model.Conf.CookieKey)
 	go func() {
 		model.InitAppearance()
+		sql.SetHanSensitive(model.Conf.Search.HanSensitive) // 须在 InitDatabase 之前设置：FTS 建表的分词器参数依赖该选项
 		sql.InitDatabase(false)
 		sql.InitHistoryDatabase(false)
 		sql.InitAssetContentDatabase(false)

kernel/mobile/kernel.go

@@ -212,6 +212,7 @@ func StartKernel(container, appDir, workspaceBaseDir, timezoneID, localIPs, lang
 	go server.Serve(false, model.Conf.CookieKey)
 	go func() {
 		model.InitAppearance()
+		sql.SetHanSensitive(model.Conf.Search.HanSensitive) // 须在 InitDatabase 之前设置：FTS 建表的分词器参数依赖该选项
 		sql.InitDatabase(false)
 		sql.InitHistoryDatabase(false)
 		sql.InitAssetContentDatabase(false)

kernel/model/conf.go

@@ -133,6 +133,10 @@ func InitConf() {
 				logging.LogErrorf("parse conf [%s] failed: %s", confPath, err)
 			} else {
 				logging.LogInfof("loaded conf [%s]", confPath)
+				if nil != Conf.Search && !bytes.Contains(data, []byte("\"hanSensitive\"")) {
+					// 旧版本配置无 hanSensitive 字段：保持既往行为（区分繁简）
+					Conf.Search.HanSensitive = true
+				}
 			}
 		}
 	}

kernel/search/hanconv.go

@@ -0,0 +1,66 @@
+// SiYuan - Refactor your thinking
+// Copyright (c) 2020-present, b3log.org
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+package search
+
+import (
+	"regexp"
+	"sort"
+	"strings"
+)
+
+// hanSimpToTrads 是 hanTradToSimp 的反向索引：简体字 -> 折叠到该简体的全部繁体字。
+var hanSimpToTrads = map[rune][]rune{}
+
+func init() {
+	for t, s := range hanTradToSimp {
+		hanSimpToTrads[s] = append(hanSimpToTrads[s], t)
+	}
+	for _, ts := range hanSimpToTrads {
+		sort.Slice(ts, func(i, j int) bool { return ts[i] < ts[j] })
+	}
+}
+
+// hanCharClass 返回与 r 繁简等价的所有字符（含 r 自身对应的简体），用于构造高亮正则。
+func hanCharClass(r rune) (ret []rune) {
+	canon := r
+	if s, ok := hanTradToSimp[r]; ok {
+		canon = s
+	}
+	ret = append(ret, canon)
+	ret = append(ret, hanSimpToTrads[canon]...)
+	return
+}
+
+// hanInsensitiveRegexp 将关键字逐字符展开为繁简等价字符类，例如 "诗经" -> "[诗詩][经經]"。
+// 仅用于搜索结果高亮；等价关系与 go-sqlite3 中 siyuan 分词器 han_insensitive 的映射表
+// 来自同一份 OpenCC TSCharacters 数据，必须保持一致。
+func hanInsensitiveRegexp(k string) string {
+	var b strings.Builder
+	for _, r := range k {
+		class := hanCharClass(r)
+		if 1 == len(class) {
+			b.WriteString(regexp.QuoteMeta(string(r)))
+			continue
+		}
+		b.WriteString("[")
+		for _, c := range class {
+			b.WriteString(regexp.QuoteMeta(string(c)))
+		}
+		b.WriteString("]")
+	}
+	return b.String()
+}